- cross-posted to:
- technology
- cross-posted to:
- technology
European Union regulators are concerned that Microsoft may be covertly controlling OpenAI as its biggest investor.
European Union regulators are concerned that Microsoft may be covertly controlling OpenAI as its biggest investor.
They’re not a majority stakeholder. They invested $13b and a lot of that wasn’t cash, it was just credits to use Azure.
That’s more money than anyone else has invested so far, but it’s only about 2 weeks revenue for Microsoft and not even close to enough to fund OpenAI.
Wether or not they’re a non profit? I dunno. OpenAI’s structure is pretty unusual.
As I understand, OpenAI’s structure is similar to Mozilla’s, so a non-profit steers a for-profit.
But there’s the massive difference that OpenAI’s for-profit takes on outside investors. That is where Microsoft has invested.
As such, control over the for-profit really isn’t in the hands of the non-profit, because obviously, they have to satisfy whoever gives them money.
In particular, it also means that unlike in Mozilla’s structure, where the for-profit is ‘neutered’ in that it can’t pay out profits to anyone, this really isn’t the case for OpenAI.
So, all in all, I really don’t feel like the non-profit part of OpenAI has any real relevance.
Right, that’s my understanding too.
Does that matter at all? Microsoft is a for-profit business, investing in a for-profit company and hoping to make a profit. Seems pretty straight forward to me.
They have to satisfy whatever terms where in the contract signed when Microsoft invested. From what I can tell, Microsoft basically just has IP rights to the software any patents. They don’t have any control.
That’s not true at all. Mozilla pays about a quarter billion dollars a year to their employees for example. They absolutely can and do user their money to pay other people.
There are strict regulations on how a non-profit can spend their money, but they absolutely are allowed (and expected) to spend all of the money they have. Generally, a non-profit is required to clearly define some purpose that benefits the community somehow (it could be helping kids with cancer, or helping elderly people get to their doctor’s appointments, or building a better Internet (Mozilla), or trying to create AGI that benefits all of humanity (OpenAI)). Whatever your purpose is, as long as the money is spent on that it’s fine.
However, the for-profit arm of OpenAI isn’t a non-profit. So it doesn’t have that same limitation.
Again, I’m not a lawyer, so I don’t know if OpenAI is legally in the clear here… but I will say that I don’t see any major problems. I also don’t think it would really be the end of the world if the EU declares they’re a for-profit enterprise. In fact I’m not sure it would change anything at all. There’s nothing wrong with running a private company. You just might have to pay a bit more tax… but since OpenAI doesn’t actually make any money… they don’t have to pay any tax anyway.
It’s also entirely possible for them to be recognised as a non-profit in some countries but not in others. Different countries, different rules. Not a big deal.
I don’t think you can co-create a supercomputer with a company and not be deeply involved with them. Combine that with the inherent power Microsoft has because of their money, and I think it’s pretty easy to see how they could exercise control indirectly.